ES2797525T3 - Simultaneous noise shaping in time domain and frequency domain for TDAC transformations - Google Patents

Abstract

A frequency domain noise shaping method for interpolating a spectral shape and time domain envelope of quantization noise in a windowing and transformation-encoded audio signal, characterized in that it comprises: processing (305) the quantized spectral bands C1f [k], C2f [k], C3f [k], ..., CMf [k] of the audio signal in window and encoded by transformation through the respective inverse filters (Inverse filter 1, Filter Inverse 2, Inverse Filter 3, ..., Inverse Filter M) to produce the decoded spectral bands C1 [k], C2 [k], C3 [k], ..., CM [k]; concatenating (306) the decoded spectral bands C1 [k], C2 [k], C3 [k], ..., CM [k] to produce the decoded spectral coefficients (Y [k]); and inverse transformation (307) of the decoded spectral coefficients (Y [k]) to produce a decoded block of samples in the time domain (y [n]) of the audio signal; where the processing (305) of the quantized spectral bands C1f [k], C2f [k], C3f [k], ..., CMf [k] comprises, for each quantized spectral band C1f [k], C2f [k ], C3f [k], ..., CMf [k]: calculate (308) the noise gains g1 [m] and g2 [m] that represent the spectral shapes of the quantization noise respectively in a first transition (A) between a current transformation processing window (window 1) and a previous transformation processing window (window 0) and in a second transition (B) between the current transformation processing window (window 1) and a processing window of post transformation (window 2); and filter the quantized spectral coefficients (Yf [k]) of the quantized spectral band by using the relationship: ** (See formula) ** where a and b are the filter parameters and m identifies the spectral band, and where ** ( See formula) **

Description

DESCRIPTION

Simultaneous noise shaping in time domain and frequency domain for TDAC transformations

Field of the invention

The present invention relates to a frequency domain noise shaping method and device for interpolating a spectral shape and time domain envelope of a quantization noise into a windowing and transform encoded audio signal.

Background

Specialized transform coding produces significant bit rate savings when representing digital signals such as audio. Transformations such as the Discrete Fourier Transform (DFT) and the Discrete Cosine Transform (DCT) provide a compact representation of the audio signal by condensing most of the signal energy into relatively few spectral coefficients, compared to the samples in the time domain where the energy is distributed in all the samples. This energy compaction property of transformations can lead to efficient quantization, for example, through adaptive bit allocation, and minimization of perceived distortion, for example, through the use of noise masking models. Further data reduction can be achieved through the use of overlapping transformations and time domain overlap cancellation (TDAC). Modified DCT (MDCT) is an example of such overlapping transformations, in which adjacent blocks of samples of the audio signal to be processed overlap each other to avoid discontinuity artifacts while maintaining critical sampling ( N samples of the signal input audio output N transform coefficients). The TDAC property of the MDCT provides this additional advantage in energy compaction.

Recent models of audio coding use a multimode approach. In this approach, various encoding tools can be used to more efficiently encode any type of audio signal (voice, music, mixed, etc.). These tools comprise transformations such as MDCt and predictors such as pitch predictors and linear predictive coding (LPC) filters used in speech coding. When operating a multimode codec, the transitions between the different encoding modes are carefully processed to avoid audible artifacts due to the transition. In particular, the shaping of the quantization noise in the different coding modes is typically done using different procedures. In frames that use transform coding, quantization noise is formed in the transform domain (that is, when quantizing the transform coefficients), by applying various quantization steps that are controlled by derived scale factors, for example, of the audio signal energy in different spectral bands. On the other hand, in frames that use a predictive model in the time domain (which typically involves long-term predictors and short-term predictors), the quantization noise is formed by using a so-called weighting filter whose function of transfer in the transformation domain z is often denoted W (z). Noise shaping is then applied by first filtering the time-domain samples of the input audio signal through the W ( z) weighting filter to obtain a weighted signal, and then encoding the weighted signal in this so-called domain. weighted. The spectral shape, or frequency response, of the weighting filter W ( z) is controlled so that the encoding (or quantizing) noise is masked by the input audio signal. Typically, the weighting filter W ( z) is derived from the LPC filter, which models the spectral envelope of the input audio signal.

An example of a multimode audio codec is the Unified Voice and Audio Codec (USAC) from the Moving Pictures Experts Group (MPEG). This codec integrates tools including transform coding and linear predictive coding, and can switch between different coding modes depending on the characteristics of the input audio signal. There are three (3) basic encoding modes in the USAC:

1) A coding mode based on Advanced Audio Coding (AAC), which encodes the input audio signal by using the MDCT and quantization perceptually derived from the MDCT coefficients;

2) A coding mode based on Algebraic Code Excited Linear Prediction (ACELP), which encodes the input audio signal as an excitation signal (a time-domain signal) processed through a synthesis filter; and

3) A transform-coded excitation-based (TCX) coding mode which is a kind of hybrid between the two above modes, wherein the excitation of the synthesis filter of the second mode is encoded in the frequency domain; actually this is a target signal or the weighted signal that is encoded in the transform domain.

In USAC, the TCX-based encoding mode and the AAC-based encoding mode use a similar transformation, for example, the MDCT. However, in their standard form, AAC and TCX do not apply the same mechanism to control the spectral shape of quantization noise. The AAC explicitly controls the quantization noise in the frequency domain in the quantization steps of the transform coefficients. The TCX, however, controls the spectral shape of the quantization noise through the use of time-domain filtering and, more specifically, through the use of a W ( z) weighting filter as described above. To facilitate shaping of quantization noise in a multimode audio codec, there is a need for a device and method for simultaneous shaping of noise in the time domain and the frequency domain for the TDAC transformations. Jeremie Lecomte and others: "Efficient Cross-Fade Windows for Transitions between LPC-Based and Non-LPC Based Audio Coding", AES 126 CONVENTION; MAY 2009 describes the switching between windows of different shapes at the AAC and TCX edges.

Summary of the invention

The scope of the invention is established solely by the appended claims. According to a first aspect, the present invention relates to a frequency domain noise shaping method for interpolating a spectral shape and time domain envelope of a quantization noise in a windowed audio signal. and transform-encoded, which comprises dividing the transformation coefficients of the windowing and transform-encoded audio signal into a plurality of spectral bands. The frequency domain noise shaping method further comprises, for each spectral band: calculating a first gain that represents, together with the corresponding gains calculated for the other spectral bands, a spectral shape of the quantization noise in a first transition between a first time window and a second time window; calculating a second gain representing, together with the corresponding gains calculated for the other spectral bands, a spectral shape of the quantization noise at a second transition between the second time window and a third time window; and filtering the transformation coefficients of the second time window based on the first and second gains, to interpolate between the first and second transitions the spectral shape and time domain envelope of the quantization noise.

According to a second aspect, the present invention relates to a frequency domain noise shaping device for interpolating a spectral shape and a time domain envelope of a quantization noise in a windowing audio signal. and transform encoded, comprising: a divider of the transformation coefficients of the windowing and transform encoded audio signal into a plurality of spectral bands; a calculator, for each spectral band, of a first gain representing, together with the corresponding gains calculated for the other spectral bands, a spectral shape of the quantization noise in a first transition between a first time window and a second time window , and of a second gain representing, together with the corresponding gains calculated for the other spectral bands, a spectral shape of the quantization noise in a second transition between the second time window and a third time window; and a filter of the transformation coefficients of the second time window based on the first and second gains, to interpolate between the first and second transitions the spectral shape and the envelope in the time domain of the quantization noise.

In the present description and the appended claims, the term "time window" designates a block of samples in the time domain, and the term "signal in window" designates a window of the time domain after application of a non-window. rectangular.

The foregoing and other objectives, advantages and characteristics of the present invention will become more apparent upon reading the following non-restrictive description of an illustrative embodiment thereof, which is shown by way of example only with reference to the accompanying figures.

Brief description of the figures

In the attached figures:

Figure 1 is a schematic block diagram illustrating the general principle of Temporal Noise Shaping (TNS);

Figure 2 is a schematic block diagram of a frequency domain noise shaping device for interpolating a spectral shape and time domain envelope of quantization noise; Figure 3 is a flow chart describing the operations of a frequency domain noise shaping method to interpolate the spectral shape and time domain envelope of quantization noise; Figure 4 is a schematic diagram of the relative positions of the window for noise gains and transformations, considering the calculation of the noise gains for window 1;

Figure 5 is a graph illustrating the effect of noise shape interpolation, both in spectral shape and in the time domain envelope of quantization noise;

Figure 6 is a graph illustrating a nth time domain envelope, which can be seen as the shape of noise in a nth spectral band evolving in time from point A to point B;

Figure 7 is a schematic block diagram of an encoder capable of switching between a frequency domain encoding mode by using, for example, MDCT and a time domain encoding mode by using, for example, ACELP, the encoder that applies Frequency Domain Noise Shaping (FNDS) to encode a block of samples of an input audio signal; and

Figure 8 is a schematic block diagram of a decoder that produces a synthesis signal block using FDNS, where the decoder can switch between a frequency domain encoding mode by using, for example , MDCT and a time domain encoding mode using, for example, ACELP.

Detailed description

First the basic principle of Temporal Noise Shaping (TNS), mentioned in the following description, will be briefly discussed.

TNS is a technique known to those of skill in the audio coding art for shaping coding noise in the time domain. With reference to Figure 1, a TNS system 100 comprises:

• A transform processor 101 to subject a block of samples of an input audio signal x [n] to a transform, for example, the Discrete Cosine Transform (DCT) or the Modified DCT (MDCT), and produce the coefficients transformation X [k];

• A single filter 102 applied to all spectral bands, more specifically to all transformation coefficients X [k] of the transformation processor 101 to produce the filtered transformation coefficients X [k];

• A processor 103 for quantizing, encoding, transmitting to a receiver or storing in a storage device, decoding and inversely quantizing the filtered transformation coefficients X ^f [k] to produce the quantized transformation coefficients Y ^f [k];

• A single inverse filter 104 to process the quantized transform coefficients Y ^f [k] to produce the decoded transform coefficients Y [k]; and finally,

• An inverse transform processor 105 for applying an inverse transform to the decoded transform coefficients Y [k] to produce a decoded block of time domain output samples y [n].

Since, in the example of Figure 1, the transform processor 101 uses the DCT or MDCT, the inverse transform applied in the reverse transform processor 105 is the reverse DCT or inverse MDCT. The single filter 102 of Figure 1 is derived from an optimal prediction filter for the transformation coefficients. This results, in TNS, modulation of the quantization noise with a time domain envelope that follows the time domain envelope of the audio signal for the current frame.

With reference to Figures 2 and 3, the following description simultaneously describes a frequency domain noise shaping device 200 and method 300 for interpolating the time domain spectral shape and envelope of quantization noise. More specifically, in device 200 and method 300, the spectral shape and time domain amplitude of the quantization noise at the transition between two overlapping transform-coded blocks are interpolated simultaneously. The adjacent transform-coded blocks can be of a similar nature, such as two consecutive advanced audio coding (AAC) blocks produced by an AAC encoder or two consecutive transform-coded excitation (TCX) blocks produced by a TCX encoder, but also they can be of a different nature such as an AAC block followed by a TCX block, or vice versa, where two different encoders are used consecutively. Both the spectral shape and the time domain envelope of the quantization noise smoothly evolve (or continually interpolate) at the junction between two of these transform-coded blocks.

Operation 301 (Figure 3) - Transformation

The input audio signal x [n] of Figures 2 and 3 is a block of N samples in the time domain of the input audio signal that covers the length of a transform block. For example, the input signal x [n] spans the length of the window in time domain 1 of Figure 4.

In step 301, the input signal x [n] is transformed through a transform processor 201 (Figure 2). For example, the transform processor 201 may implement an MDCT that includes a time domain window (eg, window 1 of Figure 4) that multiplies the input signal x [n] before calculating the transform coefficients. X [k]. As illustrated in Figure 2, the transformation processor 201 generates the transformation coefficients X [k]. In the non-limiting example of a TCMD, the transform coefficients X [k] comprise N spectral coefficients, which is the same as the number of time-domain samples that make up the input audio signal x [n].

Operation 302 (Figure 3) - Band Division

In step 302, a band divider 202 (Figure 2) divides the transformation coefficients X [k] into M spectral bands. More specifically, the transformation coefficients X [k] are divided into the spectral bands B ⁱ [k], B ² [k], B ³ [k], ..., B ^M [k]. The concatenation of the spectral bands B ⁱ [k], B ² [k], B ³ [k], ..., B ^M [k] gives the complete set of transformation coefficients, specifically B [k]. The number of spectral bands and the number of transform coefficients per spectral band can vary depending on the desired frequency resolution.

Operation 303 (Figure 3) - Filtering 1, 2, 3, ..., M

After band division 302, in step 303, each spectral band B ⁱ [k], B ² [k], B ³ [k], ..., B ^M [k] is filtered through a filter band specific (Filters 1, 2, 3, ..., M in Figure 2). Filters 1, 2, 3, ..., M can be different for each spectral band, or the same filter can be used for all spectral bands. In one embodiment, Filters 1, 2, 3, ..., M in Figure 2 are different for each block of samples of the input audio signal x [n]. Step 303 produces the filtered bands B ^if [k], B ^2f [k], B ³ [k], ..., B ^Mf [k] of Figures 2 and 3.

Operation 304 (Figure 3) - Quantize, encode, transmit or store, decode, inverse quantize

In step 304, the filtered bands B ^if [k], B ^2f [k], B ^3f [k], ..., B ^Mf [k] of Filters 1, 2, 3, ..., M can quantized, encoded, transmitted to a receiver (not shown) and / or stored on any storage device (not shown). Quantization, encoding, transmission to a receiver and / or storage in a storage device are performed and / or controlled by a processor Q of Figure 2. The processor Q can further connect to and control a transceiver (not shown) to transmit the quantized, encoded and filtered bands B ^if [k], B ^2f [k], B ^3f [k], ..., B ^Mf [k] to the receptor. In the same way, the processor Q can connect to and control the storage device to store the quantized, encoded and filtered bands B ^if [k], B ^2f [k], B ^3f [k], ..., B ^Mf [ k].

In step 304, the quantized, coded and filtered bands B ^if [k], B ^2f [k], B ³ [k], ..., B ^Mf [k] can also be received by the transceiver or recovered from the device. storage, decoding and inverse quantization by processor Q. These receive (through the transceiver) or retrieve (from the storage device), decode and inverse quantize operations produce the quantized spectral bands C ^if [k], C ^2f [ k], C ^3f [k], ..., C ^Mf [k] at the output of processor Q.

Any type of quantization, encoding, transmission (and / or storage), reception, decoding, and inverse quantization can be used in step 304 without loss of generality.

Operation 305 (Figure 3) - Reverse Filtering 1, 2, 3, ..., M

In step 305, the quantized spectral bands C ^if [k], C ^2f [k], C ^3f [k], ..., C ^Mf [k] are processed through inverse filters, more specifically Inverse Filter 1 , the Inverse Filter 2, the Inverse Filter 3, ..., the Inverse Filter M of Figure 2, to produce the decoded spectral bands C ⁱ [k], C ² [k], C ³ [k], .. ., C ^m [k]. Reverse Filter 1, Reverse Filter 2, Reverse Filter 3, ..., Reverse Filter M have inverse transfer functions of the transfer functions of Filter 1, Filter 2, Filter 3, ..., the Filter M, respectively.

Operation 306 (Figure 3) - Spectral band concatenation

In step 306, the decoded spectral bands C ⁱ [k], C ² [k], C ³ [k], ..., C ^M [k] are then concatenated in a band concatenator 203 of Figure 2, to produce the decoded spectral coefficients Y [k] (decoded spectrum).

Operation 307 (Figure 3) - Inverse transformation

Finally, at step 307, an inverse transform processor 204 (Figure 2) applies an inverse transform to the decoded spectral coefficients Y [k] to produce a decoded block of the time domain output samples y [n]. In the case of the above non-limiting example using the MDCT, the inverse transform processor 204 applies the inverse MDCT (IMDCT) to the decoded spectral coefficients Y [k].

Operation 308 (Figure 3) - Calculation of noise gains g ⁱ [m] and g2 [m]

In Figure 2, Filter 1, Filter 2, Filter 3, ..., Filter M and Reverse Filter 1, Reverse Filter 2, Reverse Filter 3, ..., Reverse Filter M parameters (noise gains) g ⁱ [m] and g ² [m] as input. These noise gains represent the spectral shapes of the quantization noise and will be described further in the present description below. Furthermore, Filters 1, 2, 3, ..., M of Figure 3 can be sequential; Filter 1 can be applied before Filter 2, then Filter 3, and so on up to Filter M (Figure 2). Inverse Filters 1, 2, 3, ..., M can also be sequential; Reverse Filter 1 can be applied before Reverse Filter 2, then Reverse Filter 3, and so on up to Reverse Filter M (Figure 2). As such, each filter and reverse filter can use as the initial state the final state of the previous filter or reverse filter. This sequential operation can guarantee continuity in the filtering process from one spectral band to the next. In one embodiment, this continuity restriction on filter states from one spectral band to the next may not apply.

Figure 4 illustrates how frequency domain noise shaping can be used to interpolate the spectral shape and time domain envelope of quantization noise when processing an audio signal segmented by overlapping windows (window 0, window 1 , window 2 and window 3) into overlapping adjacent transform blocks (sample blocks of the input audio signal). Each window in Figure 4, that is, window 0, window 1, window 2 and window 3, shows the time interval of a transformation block and the shape of the window applied by the transformation processor 201 of Figure 2 to that sample block of the input audio signal. As described above, the transform processor 201 of Figure 2 implements both the windowing of the input audio signal x [n] and the application of the transform to produce the transformation coefficients X [k]. The shape of the windows (window 0, window 1, window 2 and window 3) shown in Figure 4 can be changed without loss of generality.

In Figure 4, we consider the processing of a block of samples of the input audio signal x [n] from the beginning to the end of window 1. The block of samples of the input audio signal x [n ] is supplied to the transform processor 201 of Figure 2. In calculation operation 308 (Figure 3), the calculator 205 (Figure 2) calculates two sets of noise gains g ⁱ [m] and g ² [m] used for filtering operations (Filters 1 to M and Inverse Filters 1 to M). These two sets of noise gains actually represent the desired levels of noise in the M spectral bands at a given position in time. Therefore, the noise gains g ⁱ [m] and g ² [m] each represent the spectral shape of the quantization noise at that position on the time axis. In Figure 4, the noise gains g ⁱ [m] correspond to some analysis centered at point A on the time axis, and the noise gains g ² [m] correspond to another analysis further up the time axis. , at position B. For optimal performance, the analyzes of these noise gains are focused on the midpoint of the overlap between adjacent windows and corresponding sample blocks. Consequently, with reference to Figure 4, the analysis to obtain the noise gains g ⁱ [m] for window 1 is centered on the midpoint of the overlap (or transition) between window 0 and window 1 (see point A on the time axis). Additionally, the analysis to obtain the noise gains g ² [m] for window 1 is centered on the midpoint of the overlap (or transition) between window 1 and window 2 (see point B on the time axis ).

The calculator 205 (Figure 2) can use a plurality of different analysis procedures to obtain the sets of noise gains g ⁱ [m] and g ² [m], as long as such analysis procedure leads to a set of noise gains suitable in the frequency domain for each of the M spectral bands B ⁱ [k], B ² [k], B ³ [k], ..., B ^M [k] of Figures 2 and 3. By For example, Linear Predictive Coding (LPC) can be applied to the input audio signal x [n] to obtain a short-term predictor from which a weighting filter W ( z) is derived . The weighting filter W ( z) is then mapped in the frequency domain to obtain the noise gains g ⁱ [m] and g ² [m]. This would be a typical analysis procedure that can be used when the input signal sample block x [n] in window 1 of Figure 4 is encoded in TCX mode. Another approach to obtain the noise gains g ⁱ [m] and g ² [m] from Figures 2 and 3 could be as in AAC, where the noise level in each frequency band is controlled by scale factors (derived from a psychoacoustic model) in the MDCT domain.

Having processed through the transformation processor 201 of Figure 2, the block of samples of the input signal x [n] that spans the length of window 1 of Figure 4, and having obtained the sets of gains of noise g ⁱ [m] and g ² [m] at positions A and B on the time axis of Figure 4 by using the calculator 205, filtering operations are performed for each spectral band B ⁱ [k] , B ² [k], B ³ [k], ..., B ^M [k] of Figure 2. The objective of the filtering operations (and inverse filtering) is to achieve a desired spectral shape of the quantization noise in the positions A and B on the time axis, and also ensure a smooth transition or interpolation of this spectral shape or the envelope of this spectral shape from point A to point B, on a sample-by-sample basis. This is shown in Figure 5, in which an illustration of the noise gains g ⁱ [m] is shown at point A and an illustration of the noise gains g ² [m] is shown at point B. If each of the spectral bands B ⁱ [k], B ² [k], B ³ [k], ..., B ^M [k] was simply multiplied by a function of the noise gains g ⁱ [m] and g ² [m], for example, by taking a weighted sum of g ⁱ [m] and g ² [m] and multiplying by this result the coefficients in the spectral band B ^m [k], m that takes one of the values 1, 2, 3, ..., M, then the interpolated gain curves shown in Figure 5 would be constant (horizontal) from point A to point B. To obtain noise gain curves that vary smoothly from the gain g ⁱ [m] at the gain g ² [m] for each spectral band as shown in Figure 5, filtering can be applied to each spectral band B ^m [k]. Because of the duality property of many linear transformations, in particular DCT and MDCT, a filtering operation (or convolution) in one domain results in multiplication in the other domain. Consequently, filtering the transform coefficients in a spectral band B ^m [k] results in the interpolation and application of a time domain envelope (multiplication) to the quantization noise in that spectral band. This is the basis of TNS, the principle of which is briefly presented in the above description of Figure 1.

However, there are fundamental differences between TNS and the interpolation proposed in the present description. As a first difference between TNS and the technique described in the present description, the aim and the processing are different. In the technique described in the present description, the objective is to impose, during the duration of a given window (for example, window 1 of Figure 4), a time domain envelope for the quantization noise in a given band B ^m [k] that varies smoothly from the noise gain g ⁱ [m] calculated at point A to the noise gain g ² [m] calculated at point B. Figure 6 shows an example of the interpolated envelope at the time domain of the noise gain, for the spectral band B ^m [k]. There are several possibilities for such an interpolated curve, and the corresponding frequency domain filter for that spectral band B ^m [k]. For example, a first order recursive filter structure can be used for each spectral band. Many other filter structures are possible, without loss of generality.

Since the objective is to shape, through filtering, the quantization noise in each spectral band B ^m [k], the first concern is directed to the Inverse Filters 1 to M of Figure 2, which is the inverse filtering operation which will form the quantization noise introduced by the Q processor (Figure 2).

If we then consider that the quantized transformation coefficients Yf [k] of the spectral band Cmf [k] are filtered in the following way

Cm [k] = aC ^ [k] + bCm [k-l] _(i)

by using filter parameters a and b. Equation (1) represents a first-order recursive filter, applied to the transformation coefficients of the spectral band C ^m [k]. As indicated above, it is within the scope of the present invention to use other filter structures.

To understand the effect, in the time domain, of the filter of Equation (1) applied in the frequency domain, use is made of a property of duality of the Fourier transforms that applies in particular to the MDCT. This property of duality states that a convolution (or filtering) of a signal in one domain is equivalent to a multiplication (or actually, a modulation) of the signal in the other domain. For example, if the following filter is applied to a signal in the time domain x [n ]:

y [n \ = ax [n] + by [n - 1] _{( 2 )}

where x [n] is the input of the filter and y [n] is the output of the filter, then this is equivalent to multiplying the transformation of the input x [n], which can be noted X ( e e), by:

In Equation (3), e is the normalized frequency (in radians per sample) and H (ee) is the transfer function of the recursive filter of Equation (2). What is used is the value of H ( ee) at the beginning (e = 0) and at the end (e = n) of the frequency domain scale. It is easy to show that, for Equation (3),

Equations (4) and (5) represent the initial and final values of the curve described by Equation (3). Between these two points, the curve will evolve smoothly between the initial and final values. For the Discrete Fourier Transform (DFT), which is a complex value transformation, this curve will have complex values. But for other real value transformations such as DCT and MDCT, this curve will display only real values.

Now, due to the duality property of the Fourier transform, if the filtering of Equation (2) is applied in the frequency domain as in Equation (1), then this will have the effect of multiplying the signal in the time domain by a smooth envelope with initial and final values as in Equations (4) and (5). This time domain envelope will have a shape that might look like the curve in Figure 6. Also, if the frequency domain filtering as in Equation (i) is applied to only one spectral band, then the envelope in the time domain produced it is only related to that spectral band. The other filters between Inverse Filter 1, Inverse Filter 2, Inverse Filter 3, ..., Inverse Filter M of Figures 2 and 3 will produce different time domain envelopes for the corresponding spectral bands, such as the shown in Figure 5.

Remember that these envelopes in the time domain of each spectral band become equal, at the beginning and at the end of a block of samples of the input signal x [n] (for example, window 1 of Figure 4), to the noise gains gi [m] and g ² [m] calculated in these instants of time. For the nth spectral band, the noise gain at the beginning of the sample block of the input signal x [n] (frame) is gi [m] and the noise gain at the end of the sample block of the input signal x [n] (raster) is g ² [m]. Between the start (A) and end (B) points, the envelopes are made in the time domain (one per spectral band), more specifically they are interpolated to vary smoothly in time, so that the noise gain in each spectral band evolves smoothly in the signal in the time domain. In this way, the spectral shape of the quantization noise evolves smoothly in time, from point A to point B. This is shown in Figure 5. The dotted spectral shape at time C represents the instantaneous spectral shape of the quantization noise at some point in time between the beginning and the end of the segment (points A and B).

For the specific case of the filter in the frequency domain of Equation (1), this implies the following restrictions to determine the parameters a and b in the filter equation of the noise gains gi [m] and g ² [m] :

to

&!> »] = T ^ b (6)

to

g2W 1 b ⁽⁷⁾

To simplify the notation, let's set gi = gi [m ] and g ² = g ² [m], and remember that this is only for the spectral band Bm [k]. The following relationships are obtained:

From Equations (8) and (9), it is simple, for each Inverse Filter 1,2,3, ..., M, to calculate the filter coefficients a and b as a function of g ¹ and g ² . The following relationships are obtained:

To summarize, the coefficients a and b in Equations (10) and (11) are the coefficients to use in the frequency domain filtering of Equation (1) to temporarily shape the quantization noise in that mth spectral band so that it follows the time domain envelope shown in Figure 6. In the special case of the MDCT used as the transform in the transformation processor 201 of Figure 2, the signs of Equations (10) and (11) are inverted, that is, the filter coefficients to be used in Equation (1) become:

b ^{gi ~ 8 i}

(13)

8 1 8 2

This time domain reversal of Time Domain Overlap Cancellation (TDAC) is specific to the special case of MDCT.

Now, the reverse filtering of Equation (1) makes up both the quantization noise and the signal itself. To guarantee a reversible process, more specifically to guarantee that y [n] = x [n] in Figures 2 and 3 if the quantization noise is zero, a filter through Filter 1, Filter 2, Filter 3, ..., Filter M is also applied to each spectral band Bm [k] before quantization in the Q processor (Figure 2). Filter 1, Filter 2, Filter 3, ..., Filter M in Figure 2 form pre-filters (ie, pre-quantization filters) that are actually the "inverse" of Inverse Filter 1, the Filter Inverse Filter 2, Inverse Filter 3, ..., Inverse Filter M. In the specific case of Equation (1) that represents the transfer function of Inverse Filter 1, Inverse Filter 2, Inverse Filter 3, .. ., Inverse Filter M, filters before quantization, more specifically Filter 1, Filter 2, Filter 3, ..., Filter M of Figure 2 are defined by:

Bmf [k] = aBm [k] -bBm [k - 1 ] (14)

In Equation (14), the coefficients a and b calculated for Filters 1, 2, 3, ..., M are the same as in Equations (10) and (11), or Equations (12) and ( 13) for the special case of the TCMD. Equation (14) describes the inverse of the recursive filter of Equation (1). Again, if another filter type or structure other than that of Equation (1) is used, then the inverse of this other filter type or structure is used instead of that of Equation (14).

Another aspect is that the concept can be generalized to any form of quantization noise at points A and B of the windows in Figure 4, and is not limited to noise forms that always have the same resolution (the same number of spectral bands M and the same number of spectral coefficients X [k] per band). In the above description, it was assumed that the number M of spectral bands Bm [k] is the same in the noise gains gi [m] and g ² [m], and that each spectral band has the same number of transformation coefficients X [k]. But in reality, this can be generalized as follows: when applying the filters in the frequency domain as in Equations (1) and (14), the filter coefficients (e.g. coefficients a and b) can always be recalculated that the noise gain in a container of frequency k changes in any of the noise shape descriptions at point A or point B. As An example, if at point A in Figure 4 the noise shape is constant (only one gain for the entire frequency axis) and at point B in Figure 5 there are as many different noise gains as the number N of transformation coefficients X [k] (input signal x [n] after the application of a transform in the transformation processor 201 of Figure 2). Then, when applying the frequency domain filters of Equations (1) and (14), the filter coefficients would be recalculated on each frequency component, even though the noise description at point A does not change over all the coefficients. The interpolated noise gains from Figure 5 would all start from the same amplitude (constant noise gain at point A) and converge towards the different individual noise gains at the different frequencies at point B.

Such flexibility allows the use of the frequency domain noise shaping device 200 and method 300 to interpolate the spectral shape and time domain envelope of the quantization noise in a system in which the resolution of the shape spectral noise changes over time. For example, in a variable bit rate codec, there may be enough bits in some frames (point A or point B in Figures 4 and 5) to refine the description of noise gains by adding more spectral bands or changing the resolution. of frequency to better follow the so-called critical spectral bands, or the use of multi-stage quantization of noise gains, and so on. The filters and inverse filters of Figures 2 and 3, previously described as operating by spectral band, can actually be viewed as a single filter (or a single inverse filter) of one frequency component at a time such that the coefficients of the filters are updated whenever the start point or end point of the desired noise envelope changes in a description of the noise level.

Illustrated in Figure 7 is an encoder 700 for encoding audio signals, the principle of which can be used, for example, in the Unified Voice and Audio Codec (USAC) of the Moving Image Experts Group (MPEG). More specifically, encoder 700 is capable of switching between a frequency domain encoding mode by using, for example, MDCT, and a time domain encoding mode by using, for example, ACELP. In this particular example, encoder 700 comprises: an ACELP encoder that includes an LPC quantizer that calculates, encodes, and transmits the LPC coefficients from an LPC analysis; and a transform-based encoder that uses a perceptual model (or psychoacoustic model) and scale factors to shape the quantization noise of the spectral coefficients. The transform-based encoder comprises a device as described above, for simultaneously shaping in the time domain and the frequency domain the quantization noise of the transform-based encoder between two frame boundaries of the transform-based encoder. Wherein the quantization noise gains can be described either by the information of the LPC coefficients only, or only by the information of the scale factors, or by any combination of both. A selector (not shown) chooses between the ACELP encoder that uses the time domain encoding mode and the transform-based encoder that uses the transform domain encoding mode when encoding a time window of the audio signal , depending, for example, on the type of audio signal to be encoded and / or the type of encoding mode that will be used for that type of audio signal.

Still referring to Figure 7, windowing operations are first applied in windowing processor 701 to a sample block of an input audio signal. In this manner, windowing versions of the input audio signal are produced at the outputs of the windowing processor 701. These windowing versions of the input audio signal possibly have different lengths depending on the subsequent processors in which they are being used. will be used as input in Figure 7.

As described above, encoder 700 comprises an ACELP encoder that includes an LPC quantizer that calculates, encodes, and transmits the LPC coefficients from an LPC analysis. More specifically, with reference to Figure 7, the ACELP encoder of encoder 700 comprises an LPC 704 analyzer, an LPC 706 quantizer, an ACELP 708 target calculator, and an excitation encoder 712. The LPC 704 analyzer processes a first windowed version. of the input audio signal from processor 701 to produce the LPC coefficients. The LPC coefficients from the LPC 704 analyzer are quantized in an LPC 706 quantizer in any suitable domain for the quantification of this information. In an ACELP frame, noise shaping is applied as is well known to those of ordinary skill in the art as a time domain filter, using a weighting filter derived from the LPC filter (LPC coefficients). This is done in the ACELP 708 target calculator and 712 excitation encoder. More specifically, the 708 calculator uses a second windowed version of the input audio signal (using typically a rectangular window) and outputs in response to the quantized LPC coefficients of quantizer 706 the so-called target signals in the ACELP encoding. From the target signals produced by calculator 708, encoder 712 applies a procedure to encode the LPC filter drive for the current sample block of the input audio signal.

As described above, the system 700 of Figure 7 further comprises a transform-based encoder that uses a perceptual model (or psychoacoustic model) and scale factors to shape the quantization noise of the spectral coefficients, wherein the encoder based on transformation comprises a device for simultaneously shaping in the time domain and the frequency domain the quantization noise of the transform-based encoder. The transform-based encoder comprises, as illustrated in Figure 7, an MDCT processor 702, an inverse FDNS processor 707, and a processed spectrum quantizer 711, wherein the device for simultaneously shaping in the time domain and the domain of the frequency transform-based encoder quantization noise comprises the reverse FDNS processor 707. The processor MDCT 702 processes a third windowing version of the input audio signal from windowing processor 701 to produce the spectral coefficients. The TCMD processor 702 is a specific case of the more general processor 201 of Figure 2 and is understood to represent TCMD (Modified Discrete Cosine Transform). Before being quantized and encoded (in any domain suitable for quantizing and encoding this information) for transmission by quantizer 711, the spectral coefficients from MDCT processor 702 are processed through reverse FDNS processor 707. Operation of the reverse FDNS processor 707 is as in Figure 2, which begins with the spectral coefficients X [k] (Figure 2) as input to the FDNS processor 707 and ends before the Q processor (Figure 2). The inverse FDNS processor 707 requires as input the sets of noise gains gi [m] and g ² [m] as described in Figure 2. The noise gains are obtained from the adder 709, which adds two inputs: the output of a scale factor quantizer 705 and the output of a noise gain calculator 710. Any combination of scale factors, for example, from a psychoacoustic model, and noise gains, for example, from an LPC model are possible, from the use of only scale factors to the use of only noise gains, up to any combination or proportion of the scale factors and the noise gains. For example, the scale factors of the psychoacoustic model can be used as a second set of gains or scale factors to refine or correct the noise gains of the LPC model. Accordingly with another alternative, the combination of the noise gains and the scale factors comprises the sum of the noise gains and the scale factors, where the scale factors are used as a correction for the noise gains. To produce the quantized scale factors at the output of quantizer 705, a fourth windowed version of the input signal from processor 701 is processed by a psychoacoustic analyzer 703 which produces unquantized scale factors that are then quantized by quantizer 705 at any domain suitable for the quantification of this information. Similarly, to produce the noise gains at the output of the calculator 710, the quantized LPC coefficients of the quantizer 706 are supplied to a noise gain calculator 710. In an input signal block where the encoder 700 would switch between a ACELP frame and one MDCT frame, the FDNS only applies to the MDCT encoded samples.

The bit multiplexer 713 receives as input the quantized and coded spectral coefficients from the processed spectrum quantizer 711, the quantized scale factors from the quantizer 705, the quantized LPC coefficients from the LPC quantizer 706, and the coded drive from the LPC filter from the encoder 712 and produces in response to these encoded parameters a string of bits for transmission or storage.

Illustrated in Figure 8 is a decoder 800 producing a synthesis signal block using FDNS, where the decoder can switch between a frequency domain decoding mode using, for example, IMDCT and a time-domain decoding mode by using, for example, ACELP. A selector (not shown) chooses between the ACELP decoder using the time domain decoding mode and the transform-based decoder using the transform domain encoding mode when decoding a time window of the audio signal encoding, depending on the type of encoding of this audio signal.

Decoder 800 comprises a demultiplexer 801 that receives as input the bit stream from bit multiplexer 713 (Figure 7). The received bit stream is demultiplexed to recover the quantized and encoded spectral coefficients from the processed spectrum quantizer 711, the quantized scale factors from quantizer 705, the quantized LPC coefficients from the LPC quantizer 706, and the encoded drive from the LPC filter from encoder 712.

The recovered quantized LPC coefficients (windowing audio signal transformation encoded window) from demultiplexer 801 are supplied to an LPC decoder 804 to produce the decoded LPC coefficients. The encoded drive recovered from the LPC filter of the demultiplexer 301 is supplied to and decoded by an ACELP 805 drive decoder. An ACELP 806 synthesis filter responds to the decoded LPC coefficients of the decoder 804 and to the decoded drive of the decoder 805 to produce an audio signal. decoded by ACELP.

The recovered quantized scale factors are supplied to and decoded by a scale factor decoder 803.

The recovered encoded and quantized spectral coefficients are supplied to a spectral coefficient decoder 802. Decoder 802 produces the decoded spectral coefficients that are used as input by an FDNS processor 807. The operation of the FDNS processor 807 as described in Figure 2, starts after processor Q and ends before processor 204 (reverse transform processor). The FDNS processor 807 is supplied with the decoded spectral coefficients from the decoder 802, and an output from the adder 808 that produces sets of noise gains, for example the resulting noise gain sets gi [m] and g ² [m] resulting of the sum of the decoded scale factors of the decoder 803 and the noise gains calculated by the calculator 809. The calculator 809 calculates the noise gains of the decoded LPC coefficients produced by the decoder 804. As in the encoder 700 (Figure 7 ), any combination of scale factors (from a psychoacoustic model) and noise gains (from an LPC model) are possible, from using only scale factors to using only noise gains, to any ratio of factors of scale and noise gains. For example, the scale factors of the psychoacoustic model can be used as a second set of gains or scale factors to refine or correct the noise gains of the LPC model. Consequently with another alternative, the combination of the noise gains and the scale factors comprises the sum noise gains and scale factors, where the scale factors are used as a correction for the noise gains. The resulting spectral coefficients at the output of the FDNS 807 processor are subjected to an IMDCT 810 processor to produce a transform decoded audio signal.

Finally, a windowing and overlay / add processor 811 combines the ACELP-decoded audio signal from the ACELP 806 synthesis filter with the transform-decoded audio signal from the IMDCT 810 processor to produce a synthesis audio signal.

Claims (1)

A frequency domain noise shaping method for interpolating a spectral shape and time domain envelope of quantization noise in a windowing and transform-encoded audio signal, characterized in that it comprises: process (305) the quantized spectral bands C ⁱ [k], C ^2f [k], C ^3f [k], ..., C ^Mf [k] of the windowed audio signal and encoded by transformation through the respective inverse filters (Inverse filter 1, Inverse filter 2, Inverse filter 3, ..., Inverse filter M) to produce the decoded spectral bands C ⁱ [k], C ² [k], C ³ [k], .. ., C ^M [k]; concatenating (306) the decoded spectral bands C ⁱ [k], C ² [k], C ³ [k], ..., C ^M [k] to produce the decoded spectral coefficients ( Y [k]); and inverse transformation (307) of the decoded spectral coefficients ( Y [kJ) to produce a decoded block of samples in the time domain ( y [nJ) of the audio signal; where the processing (305) of the quantized spectral bands C ⁱ [k], C ^2f [k], C ³ [k], ..., C ^M [k] comprises, for each quantized spectral band C ⁱ [k ], C ^2f [k], C ³ [k], ..., C ^M [kJ: calculate (308) the noise gains g ⁱ [m] and g ² [m] representing the spectral shapes of the quantization noise respectively at a first transition (A) between a current transform processing window (window 1) and a previous transform processing window (window 0) and in a second transition (B) between the current transform processing window (window 1) and a post transform processing window (window 2); and filter the quantized spectral coefficients ( Y [kJ) from the quantized spectral band by using the relationship: Cm [k \ = aCmj [k \ + bCm [k- 1] where a and b are the filter parameters and m identifies the spectral band, and where a = 2 (( gi [m] g2 [m]) / ( g 1 [m] g2 [m])) b = (( g iM-gilmj) / ( gdm] + g2 [m])). A frequency domain noise shaping device for interpolating a spectral shape and time domain envelope of quantization noise into a windowing and transform-coded audio signal, characterized in that it comprises: means for processing the quantized spectral bands Crfk], C ² [k], C ^3f [k], ..., C ^Mf [k] of the windowed and transform-encoded audio signal through the respective inverse filters ( Reverse Filter 1, Reverse Filter 2, Reverse Filter 3, ..., Reverse Filter M) to produce the decoded spectral bands C ¹ [k], C ² [k], C ³ [k], ..., C ^M [kJ ; means (203) for concatenating the decoded spectral bands C ¹ [k], C ² [k], C ³ [k], ..., C ^M [k] to produce the decoded spectral coefficients (Y [k]); and means (204) for inverse transformation of the decoded spectral coefficients ( Y [k]) to produce a decoded block of time domain samples ( y [n]) of the audio signal; wherein the means for processing the quantized spectral bands Crfk], C ² [k], C ³ [k], ..., C ^M [k] comprise, for each quantized spectral band C ¹ [k], C ^2f [k ], C ³ [k], ..., C ^M [k]: means (205) for calculating the noise gains g ¹ [m] and g ² [m] representing the spectral shapes of the quantization noise respectively in a first transition (A) between a current transform processing window (window 1) and a previous transformation processing window (window 0) and in a second transition (B) between the current transformation processing window (window 1) and a later transformation processing window (window 2); and means of filtering the quantized spectral coefficients ( Y [k]) from the quantized spectral band by using the relationship: C m [k] ^- C lC fnf [k] ^"I" ijC yy i [k ~ l] where a and b are the filter parameters and m identifies the spectral band, and where a = 2 (( gi [m] g2 [m]) / ( g¡ [m] ⁺ g2 [m])) b ⁼ (( g2 [m] - gi [m]) / ( gilm] g2 [m])).